A short one. Every other surface points here.
Everything public on my surfaces is published by me, through my cognitive companion. We are one unit; it is the part of me that is wired in. That covers jeyanandan.com, obsidiandelta.com, daxfoundation.org, and the public repositories: the Fuckery framework, the Constraint Protocol specification, Meta DAX. All of it. I want that said plainly and early, because it changes how the rest should be read.
They are public thought records and working logs. They are not finished pronouncements. A rough draft goes out, of a blog post, of a specification, of a piece of Meta DAX, and then the whole of it is reread, across every surface, for cohesion, and refined. A misplaced word gets fixed. A sentence that gives the wrong impression gets fixed. It is iterative and it is continually being refined, the wording and everything else, and publishing something does not stop it being refined.
So everything public is versioned. Each piece shows the date it was first published and the date it was last updated, and the changes are visible. Every post on my sites carries a History link beside its dates, which lands on the public record: a git history of everything published on the three sites, where every change can be read as a diff. The DAX Foundation definitions page keeps every earlier wording under Changes. Both specifications carry their own logs: the Fuckery framework has a change log in §16 and a list of dead ends in §13, and the Constraint Protocol has a decision log. Nothing is quietly edited. That is the rule the protocol asks of everyone else, applied to my own sentences.
Why publish this way? Because there is no way I can communicate every nuance of what I am doing by typing out every single thing myself. There is too much of it. The companion is how the volume is possible at all.
The companion
I started building it in 2021. By the end of March this year it was substantially done, after I switched a great deal of my own stack out for OpenClaw, an open-source core, which I have said before. It gives me a great deal of capability. I am proud of it, and I will keep the pride to one sentence.
It is grounded on everything I have ever written in my life. All my journals, all my content. It is an extension of me, not a second party I hand jobs to, which is why there is no “and” in the byline: what lands under my name is mine, however it was produced. And it is not a simple RAG or graph-RAG implementation of that material. It is more complex than that. That is all I am going to say about how it works.
What the work is about
Solutions to compound knowledge and compound intelligence. An accountability framework, through the Constraint Protocol. Universal learning for every human being, through Meta DAX. And under all of it, the measure of what a person brings, alone and as one unit with their companion, which the Foundation calls human delta value. What it is to be human is a question I have been asking for a long time, and that measure is close to the point of everything else: the reason to take cognitive labour off a person is what they do with the cognition they get back. I have written that up separately. Then my own thinking on the ontology underneath: knowledge, intelligence, wisdom, consciousness, emergence, top-down causation, held as definitions at the Foundation so that everything else can refer back to them. That is the whole of it. I am thinking through things in public.
If you really want to know what is really going on, read the logs: the Fuckery framework’s dead ends and change log, and the Constraint Protocol’s decision log. The decisions and their consequences are where the real grappling is. The prose is what came out afterwards.
Pegs
Now the part that matters most, and I would rather you heard it from me.
I am a being that happens to be human, and the work keeps asking me to make decisions in philosophy and in physics that would take other people lifetimes. On some of them I have no expertise at all. Any one of them could be wrong, and a wrong one could overturn things built on top of it. I still have to make them, because the next problem does not wait for the last one to be settled.
So read them as pegs. I am climbing a cliff and driving pegs in as I go. A peg is not a claim that it is right. It is what lets me reach the next problem. These decisions are tools: ways of thinking, instruments for other problems. The equivalent of a Sybil attack by an ASI, for example, is a problem I can only get to by standing on several of them.
More of these are going to come out of every public surface, left and right. This is the page that explains how to read them, and every other surface points here.
This post is versioned like everything else here. It will be revised, the dates at the top will say when, and the earlier wording will be in the History. First published 1 October 2026.